Earlier this week I had the pleasure of undergoing an outpatient medical procedure in one of Seattle’s most highly regarded medical centers. Without getting into unpleasant details, let’s just say it was one of those screening exams we’re all supposed to do every 5 to 10 years beyond a certain age and that it involves threading a very long tube (scope) where the sun doesn’t shine. Fortunately, the procedure is usually done with the patient under conscious sedation. What that means is that I don’t really remember a thing about it and that’s just fine with me. However, I do have perfect recollection of everything that transpired up until the time I felt the warmth of those sedating medications shoot through my IV. That’s what I want to share with my readers.
As I said, this took place in one of Seattle’s premier medical centers. They have an enterprise information system and an electronic medical record for their hospital and clinics. However, most of the physicians (especially those in the medical center’s satellite clinics) complain bitterly that the system is too hard to use, too inflexible, and doesn’t meet the needs of physician work-flow; despite the fact that the medical group has invested tens of millions of dollars in this solution.
As I sat in my hospital gown being interviewed and prepped by the IV sedation nurse, I was struck by how many pieces of paper I spotted. There must have been 10 or 15 forms scattered across the room. There were forms for my medical history, allergies, medications, consent, discharge, operative report, anesthesia, and a whole bunch more. “Why”, I asked, “were there so many pieces of paper in an institution that was supposed to have gone electronic?” My IV nurse replied that the systems in place were too hard to use, often unreliable, and that many of the physicians trusted their paper forms more than they did putting vital patient information into the computer.
Dr. Michael Wilkes who writes for the Sacramento Bee may have hit the nail on the head in a recent column entitled “Inside Medicine: So far, electronic records don’t help patients much”. But the point he’s really making is that so far electronic medical records aren’t helping patients or doctors very much. He writes, “Today, a flat-screen computer sits between the doctor and the patient -- just as a fence divides two neighbors. My students and residents -- like doctors around the country -- are slaves to the computer and electronic medical records. If you've not had the experience of sitting across from your doctor as she or he types your medical history into the computer, then just wait”.
Dr. Wilkes also takes a well deserved pot shot at systems that are too expensive, proprietary, and don’t talk to one another. He goes on to say that many of the systems in use today rely on templates for data entry. While this speeds work-flow and helps validate billing codes it also has “a potential for electronic forgery and dishonesty that allows for increased billing, and quick note production, but may do nothing to improve patient care. In fact, it may hinder care and could lead to major problems”.
But before we place blame on physicians for being old fashioned or technophobic, let’s look at the root cause of this push-back on electronic records. How many times have I pointed out that physicians meet the very definition of information worker? They are the ultimate information workers. Have I not repeatedly stated that they deserve the very best tools and technologies our industry can deliver? They must have solutions that are entirely intuitive and require very little training. These solutions must offer every means of data input; from digital ink and voice, to “point and click”, to audio and video capture, and more. These solutions must extend across an entire range of devices to provide information at the point of care and everywhere in between. They absolutely, positively must be mobile. They have to be interoperable, and they should cost a whole lot less than most of the solutions on the market today.
We have the technologies necessary to meet these work-flow requirements. Why aren’t we seeing better solutions on the market? I implore all industry solution vendors to work together to figure this out. If we don’t, some government bureaucracy will do it for us and I guarantee that not one of us, technologists or clinicians, will be very happy with what we get.
What do you think? We’d love to hear from you.
Bill Crounse, MD Healthcare Industry Director Microsoft
Your comments, and those of Dr. Wilkes, are right on the money. Here we are with a healthcare system that can offer miracles through amazing technology - drug-eluting stents, endovascular treatment for brain aneurysms, joint replacements, many more - but we can't deliver efficient, cost-effective, user-friendly, customer-oriented care. The clinical transformation through information technology that people talk about is nowhere to be seen. As you observed, todays EMR frequently just frustrates doctors, and often slows us down.
Like you, I'm very interested in this area. You may have come across the IT system we've developed here at UCLA in Neurosurgery. This application integrates with disparate multi-vendor and home-grown hospital systems to pull the key pieces of clinically-critical information together, then displays it in a faily intuitive dashboard interface, and delivers it to the attendings, residents, nurses, nurse-practioners, and physician assistants, discharge planners, utilization review nurses, etc. The information can be delivered to static desktop workstations, laptops on carts, Tablet PCs, PDAs, and Smartphones. For group ICU data/radiology rounds it can be displayed on large wall-mounted plasma screens. It can be delivered on paper as a rounding-list with latest labs and vitals that the resident teams managing a 40-patient service can generate in 90 seconds (saves that pre-rounding intern 60-90 minutes/day). It can be delivered as a ward patient-list to the Charge Nurse - saving the nurses 15-20 minutes each at end-of-shift.
We have helped the quality and speed of some work processes - but the going is remarkably slow. There is a great shortage of people who understand both clinical workflow and IT. The best successes that we have had involved recognizing opportunities for automating "clerical - level" work, and delivering a data product that looks very much like the manually-generated physician document, list or note. In such cases we recognized the painful slow mindless tasks that could be automated. Once you start looking for opportunities like this, the hospital becomes a very target rich environment. The key steps involve acquiring all the clinically-important data, and putting it into a properly structured database, and then having a front-lines nurse or doctor foot-soldier-type understanding of the clinical workflow.
Neil Martin, MD
Chief, Division of Neurosurgery
Thanks for taking time to comment. I especially honed in on the last few lines where you said;
"The best successes that we have had involved recognizing opportunities for automating "clerical - level" work, and delivering a data product that looks very much like the manually-generated physician document, list or note. In such cases we recognized the painful slow mindless tasks that could be automated. Once you start looking for opportunities like this, the hospital becomes a very target rich environment. The key steps involve acquiring all the clinically-important data, and putting it into a properly structured database, and then having a front-lines nurse or doctor foot-soldier-type understanding of the clinical workflow."
This resonates very well with my previous Blog posting entitled "Big Healthcare Savings from Surprisingly Simple Solutions". Just as you say, a quick survey of almost any hospital or ambulatory patient care setting will reveal a myriad of paper-based processes that can be computerized and automated; often with lower-cost, readily available, easy to learn and use commodity software solutions.
Congratulations on your fine work at UCLA and thank you again for writing.
Bill Crounse, MD Healthcare Industry Director Microsoft
The experience you had is typical of most hospitals I've encountered in my 25+ years working from the vendor side and by working with customers directly I heard their frustrations load and clear when it comes to poor integration amongst the various systems in most hospitals. While the solutions we developed to provide better integration were just a small part of the total hospital system they did provide better throughput and more accurate records for their users. Much more needs to be done to improve efficiency and eliminate or digitize paper records and hopefully we'll be able to develop new methods for this effort with tools such as Azyxxi and other programs.
I also work as a consultant installing and customizing electronic medical records and serve on a number of state-wide medical associations and it is very frustrating to see how expensive and fragmented HealthCare IT is. The cost to implement and support multiple systems is staggering and although Microsoft’s new product meets an existing soft spot it isn’t the long term solution. We need systems that work across entire regions not simply within a hospital. Here in Seattle for example 3 of the 4 largest systems have the same vendor but they can't share patient data due to customizations.
Cedars-Sinai actually pulled their EMR system out a few years ago and at least six other hospitals have as well in the past few years so the technology itself isn’t the solution. In 2005, Connecting for Health--a public-private cooperative of hospitals, health plans, employers and government agencies--found that persuading doctors in small- to medium-sized practices to adopt electronic medical records required offering bonuses of up to 10 percent of the doctors' annual income.
We are starting to see regional coalitions of people who are trying to publish quality and patient safety data but that is also fragmented. One example in Washington is a five county coalition - the Puget Sound Health Alliance (I serve on their Health Information Technology Board) but it will use different data then the State Hospital Association will. Simply having the data and the technology is just the first step in meeting the goal of patient centered care.
It is also mistake to assume that government programs result in more expensive products or services. Medicare has a 2% overhead versus the typical 15% at private insurance companies. I realize that I might be unique in this viewpoint since most of us make a living by selling products that we know aren’t integrated but there is in already a very robust national electronic medical record system in use at all of the VA hospitals and clinics and it is free! (not counting implementation and support) They were the only system in Katrina able to provide seamless care to their patients. It actually has the same name as Windows new operating system but it has been out since 1996 VISTA. http://www1.va.gov/CPRSdemo/ This is the demo but the full program is actually free to use. It doesn't have anyone marketing or much support yet though.
Not many people realize the dramatic turn-around that has occurred at the VA over the past 10 years. Now the VA sets the benchmark for high quality cost effective care now. "Outside experts agree that the VHA has become an industry leader in its safety and quality measures. Dr. Donald M. Berwick, president of the Institute for Health Care Improvement and one of the nation's top health-care quality experts, praises the VHA's information technology as "spectacular." The venerable Institute of Medicine notes that the VHA's "integrated health information system, including its framework for using performance measures to improve quality, is considered one of the best in the nation." http://www.washingtonmonthly.com/features/2005/0501.longman.html
The National Committee for Quality Assurance today ranks health-care plans on 17 different performance measures and who do you suppose this year's winner is: Johns Hopkins? Mayo Clinic? Massachusetts General? Nope. In every single category, the VHA system outperforms the highest rated non-VHA hospitals. In 2003 the prestigious New England Journal of Medicine published a study that compared veterans health facilities on 11 measures of quality with fee-for-service Medicare. On all 11 measures, the quality of care in veterans facilities proved to be "significantly better."
The real challenge isn't a technological one at all actually. Clearly we could roll the VA system out across the country and have an integrated system that is cost effective, easy to support and provides care of the highest quality. We could then use our talents and companies to find ways to add value added services or products that actually improve patient care. The largest challenges will not be technological but cultural and organizational one’s when we start to put patients at the center of their care (the VA system for example lets patients document in their own charts and via a web portal give other doctors outside of the VA access to their entire record). Technology is just one tool to accomplish that.
So “when it comes to health care, it's a government bureaucracy that's setting the standard for maintaining best practices while reducing costs, and it's the private sector that's lagging in quality. That unexpected reality needs examining if we're to have any hope of understanding what's wrong with America's health-care system and how to fix it. It turns out that precisely because the VHA is a big, government-run system that has nearly a lifetime relationship with its patients, it has incentives for investing in quality and keeping its patients well--incentives that are lacking in for-profit medicine.”
BTW: I am disappointed that an earlier post that also articulated the role that government could play was never posted. If we don’t hear from all voices or bring all of the players to the table (government purchases over 40% of all health care in the US). How can we hope to transform it?
Thanks for writing and sharing your thoughts. While I'm not ready to concede that rolling out the VA system to every doctor in America (public and private sector) is the only way to solve our healthcare IT woes, I do greatly appreciate the fine work being done by the VA. There is no question they are now reaping the rewards that come from a fully integrated and interoperable electronic medical record system for their patients. There is no question that patients are getting higher quality, safer, more satisfying and efficient care at the VA these days.
You are also correct on the need for training, organizational and cultural change management. We have all the technology we need to solve what ails the "system". I would add that we also need properly paced incentives to motivate the appropriate changes.
Thanks again for your thoughtful contribution. Also, if a previous post was lost, it wasn't because it was deleted on this end.
Bill Crounse, MD Healthcare Industry Director Microsoft
Saw your blog post re the lack of functional IT implementations & I will bite.
Microsoft has been one of the reasons why medical informatics in the UK is driving off a cliff. The oversold promotion of "consultancies" like Accenture who were not up to the task & the blind promotion of large firms over people who know the field is a case in point. Bill Gates did clinicians no favours when he dropped in to see Tony Blair at Downing Street in 2002.
I have been developing solutions for clinical use since 1996. I have not had phone calls returned from Microsoft. I have spoken to numerous people at HP, what used to be Compaq, IBM & numerous other organisations to garner support / access frameworks etc, with pretty poor results. Lets just say that when the big boys have the lobbying bases covered, there are no contracts or funding available for software that actually works.
What we do end up with are designed by numbers software that is so cumbersome to use that it actively discourages people from trying the next product.
I speak as someone who developed an EPR accessible over the internet in 1998 with provision for the full set of multimedia records. And further projects since, including one I am currently keeping alive of integrated voice recognition on PDA's.
I have stumbled upon your blog as I once again research EMRs. Our 30 provider/150 employee group does not have an EMR yet. There are several reasons:
1. Until HL7 is replaced with some new kind of XML messaging system, it is too complex and expensive for us to develop rigid ORDER/RESULT based interfaces with the hundreds of external sources of information that goes into your health care record. The initial generations of systems that allow us to receive "unsolicited" results are just coming to market. The PDF model of scanned documents is more or less useless for any purpose other than reducing storage space to the server rooms.
2. Until a company like Microsoft or Google (dig) markets an 800 lb gorilla then we are reluctant to buy in. My partners would welcome Microsoft Office MD - we would feel a bit more secure in our investment. The vast majority of systems are small and propietary as you say, and will be out of business or merged out before we retire - sometime in the next 20 years. The current ugly gorilla is probably GE, which is not really in this field, to them we are just a business to be fleeced, they are far to slow and far to good at making users completely dependent upon their poor service.
3. The current administration makes fanfare of throwing millions on EMR projects, but the Billions needed will have to come out of Doctor's pockets. Those pockets are not lined with the "media millions" but being picked clean by reduced Medicare funding. Until there is a way to make sure our fees are not going to erode by 20% next year at the whim of Congress, then who in their right mind would invest 10's of thousands of dollars in a system that does NOT actually increase productivity regardless of what the vendors show in their rather pathetic ROI calculators. I can cite from experience one of our practices replace a file clerk and saved 24,000 but then added and IT support person and contracts for network etc costing 41,000 anually.
4. Unless someone finally recognizes that physicians must relate to all the information they get in a more random fashion, then the poorly conceived interfaces of current EMRs just won't work (hence your witnessed compliants about it being too hard to use). Again, Medicare forces certail elements in a note to get paid. Computers can generate tons of such useless notes to justify charges, but that is NOT using an EMR to do healthcare.
I think the model needs to be a picture of a body - maybe even your own picture. I should be able to click on a part, and get a visual respesentation showing me all the information about it. Say a patient walks in with chest pain. I click on his chest, and I zoom in on the heart. I see then folders with lab tests, imaging studies, consultant reports, etc. I can quickly click on the imaging folder and have a chronological listing of all the tests linked to the heart (is this sounding more like an xml database?). It would take me seconds to find out all the health information related to the problem. Right now, it takes more, because we have to tab here and there, maybe those documents were scanned, then I have to open Adobe and actually read the pages, which may or may not be the right ones..... Also, with current EMRs we are generally forced to look at only one data element, I need to have all related data elements in a menubar across the side of a web page for me to click on at random.
I guess I'll just keep looking...
Bill, why not focus on the end results, did the multitude of forms effect the quality of your procedure?
Did they cover all the correct bases of quality for you?
I have personnaly worked with over 350 hospitals across the country and have had to deal with the financial and supply chain management systems, I have found that STANDARDIZATION of systems still does not work for hospitals. Each facility big and small is so unique and different.
Maybe we need to think of Customization Versus Standardization which works well for a firm in the supply value analysis business. Meaning, these hospitals are so unique that they should have customized systems that work to their unique operations, space and geographics but then develop the STANDARDIZED EXPORT or SHARE for the data that needs to be shared. More or less having the IT Vendors only have to conform on what will be shared and not have to reinvent or totally change their systems to try to share data.
Its like asking hospitals to standardize on one pacemaker when they have to look at the unique medical indications for the patient first then determine what pacemaker to implant which could cost the hospital between $3000 to $25,000.
My point is, healthcare is CUSTOMIZED and not a good candidate for Standardizing IT, we are just too far gone and too imbedded with current systems with major investments made. Plus, they don't have huge dollars to invest in IT when they need to keep up on their Clinical, Support Service equipment, physical plant, etc.
Thanks for writing. I agree that healthcare is a custom business. In fact, one of the reasons that so many attempts to automate clinical workflow fail is because they do not take into account the localization and customization of care. Having said that, I am not advocating standardization so much as I am advocating commoditization. We can get rid of paper and replace it with forms that are both electronic and highly flexible; allowing end users to customize the forms according to their unique business or clinical requirements. Data can be captured and integrated into the longitudinal record. Furthermore, this can be done with highly flexible, commodity software. We have released a blueprint to help organizations achieve this goal. It provides worldwide best practices for building SOA web services for the healthcare industry. For more information, please visit http://msdn.microsoft.com/health.
Bill Crounse, MD
Worldwide Health Director
Taps his fingers waiting for comment........
I haven't been ignoring you. I asked one of my colleagues in the UK to respond, but obviously that didn't happen.
I would certainly encourage you to get involved in ms-hug (www.mshug.org). We also have a partner program for solution vendors who build on our platform. If you want information on that, let me know. In addition, we have published the Connected Health Framework; an architectural guideline and blueprint to help healthcare organizations and vendors build world class web services solutions for the industry. You will find that information here: http://msdn.microsoft.com/health.
While I don't agree with you comments, I want to thank you for sharing your perspective. I wish you the very best in your endeavors.